GLOBAL ENVIRONMENT FACILITY
PROJECT CONCEPT NOTE FOR PIPELINE ENTRY AND PDF-B REQUEST
Project:
Western Kenya Integrated Ecosystem Management Project
Country Eligibility:
Kenya ratified the UNFCCC in 1994, the CBD in 1994, and the CCD
in 1997.
Focal Area:
Multifocal and addressing land degradation issues
Operational Programme:
OP 12 Integrated Ecosystem Management
Project Cost:
US$ 5.125 million (5 year period)
Financing:
IDA
GEF
GoK
ICRAF
EU
Implementing Agency:
The World Bank
Executing Agency:
Kenya Agricultural Research Institute (KARI) and the International
Center for Research in Agro-Forestry (ICRAF) as the lead agencies
in partnership with specific NGOs, local community organizations,
and Kenya Forestry Research Institute (KEFRI).
PDF-B Block B Request:
GEF
$ 200,000
PDF Co-financing:
GoK/KARI
ICRAF
IDA (NARP II)
$ 133,500
$ 55,500
$ 116,000
Block A Grant Awarded:
None
$ 0.6 million (NARP II, NARP III)
$ 2.9 million
$ 0.475 million (in cash/kind)
$ 0.5 million (in cash/kind)
$ 0.65 million
1
1. PROJECT OBJECTIVES AND SUMMARY
Project:
Geographic
Focus:
Objectives:
OUTPUTS:
Western Kenya Integrated Ecosystem Management Project
Kenyan portion of the Lake Victoria Watershed

Enhance potential to sequester additional above- and below-ground carbon in
the project area, and develop scientifically sound and cost effective
procedures and protocols to measure, monitor and validate above- and belowground carbon sinks in different land-use systems;

Promote agroforestry and other improved land management activities in
upland areas to rehabilitate degraded lands, improve erosion and sediment
control, and reduce nutrient delivery to Lake Victoria from agricultural
activities;

Improve on-farm and off-farm biodiversity through agrobiodiversity and
decreased pressure on “natural habitats” (remnant forests, riparian areas,
wetlands, etc.);

Reduce rural poverty and improve food security by introducing new land
management technologies to increase yields, and new and improved valueadded cropping systems.

Improve capacity for local communities, farmer associations, national, and
international institutions to identify opportunities, and formulate and
implement policies in support of IEM approaches, combining local and
global benefits;

Examine the synergy between sustainable agricultural development and global
environmental benefits such as mitigation of GHG accumulation in the
atmosphere, forestry, biodiversity loss, and degradation of international
waters;

Assessment of carbon sinks, sources, and reservoirs in representative land
use systems, scientifically sound estimates of potential gains and rates of
change, development of procedures to monitor, assess, and implement total
green house gas accounting in tropical rural landscapes;

Rehabilitated rural landscapes with improved stocks of soil organic matter,
improved soil fertility, and improved erosion and sediment control;

Reduced sediment and nutrient loads to Lake Victoria

Increased incidence of endemic species in agricultural landscapes, and
reduced pressure on protected areas and other biodiversity-rich areas of
global importance;

Improved rural livelihoods and economic performance of local, small-scale
farming systems through yield increases and value added cropping;

Promotion of IEM approaches and increased community involvement in IEM
activities (assessment, planning, decision making, implementation,
evaluation, etc.);

Strengthened institutional capacity to identify opportunities, and formulate
and implement policies in support of IEM approaches;

Results and lessons learned disseminated in the country and internationally.
2
2. PROJECT BACKGROUND AND DEVELOPMENT STRATEGY
“Conservation is not the goal, but rather the result of good farming practices” –C. E. Kellogg
Food insecurity and land degradation are most acute in sub-Saharan Africa where per capita food
production continues to decrease, in contrast to sustained increases in other parts of the world. This
problem is worsened by macroeconomic and financial policies that have resulted in decreased
fertilizer availability and use by small-scale farmers (FAO, 2000). The highlands of western Kenya
are home to 12 million people, or 40% of the country’s population, but occupy only 15% of the land
area. These lands have high agricultural potential, yet recent soil degradation has led to incidences of
abject poverty on the order of 30 to 50% of the rural households (Central Bureau of Statistics, 1998).
Low yields and increasing population in western Kenya have caused more marginal lands to be
brought into production and degradation of remaining natural forests, resulting in still greater rates of
land degradation, habitat destruction, and biodiversity losses. Current smallholder practices are no
longer adequate to meet food needs or maintain the resource base, a situation that must be reversed if
social and environmental disasters are to be averted. Poverty reduction, environmental quality, and
sustainable agriculture are intricately linked in the region. Reversing the downward trend will require
a sustained annual growth rate in agricultural production of 4% (Cleaver and Schreiber, 1994). This
target can only be achieved through agricultural intensification. However, the intensification must be
achieved in a manner which enhances soil fertility, the key to which, in many cases, is increasing soil
organic matter stocks. It is only through integrated management of the landscape (integrating
management of production and environmental service functions) that the dual issues of land
degradation and rural poverty reduction will be addressed, and local, national, and global
environmental benefits will be achieved.
Traditional land management in western Kenya has relied on fallowing of unproductive fields to
restore fertility and decrease pest problems. The rapid increase of population density makes this
practice untenable and has led to wide scale abandonment of fallowing. However, more people do not
necessarily lead to degradation, but the scale of increase in population pressure in W. Kenya has been
enormous in the past half century. High rural population growth1 coupled with stagnating urban job
growth has accelerated the search for new agricultural land, resulting in a high rate of woodlands,
forests, and wetlands having been converted into agriculture. The reasons for a lack of vibrancy in
urban growth are many and include several that are affected by policy or politicians: poor
infrastructure, high costs of capital, and political interference in the private sector. Locally, there has
been little restriction on encroachment onto steep slopes, wetlands, or forests, despite the existence in
some cases of laws and regulations against such practices.
Intensification of land use is necessary to achieve farming systems that are more sustainable than what
is available today. Once settled on their own land, farmer management of land is greatly affected by
the potential rewards of different enterprises and practices. Increased profitability of agriculture
increases the incentives for landowners to invest in their land, with likely implications that less
degradation will occur on their land and they will have less incentive to leave smallholdings in search
of larger ones. Experiences from Central Kenya, where there is evidence of high productivity, high
profits, and good land management, are supportive of this relationship. The government has made
some good reforms to enable markets to function better, but the agricultural sector is still plagued by
poor management of parastatals (e.g. cooperatives), politicization of key commodity sectors, and
inadequate maintenance and expansion of infrastructure. The problems of infrastructure of course
affect the credit and input side of the ledger as well. Credit is a serious problem for the small farmer,
but this is not unique to Western Kenya, being similarly constraining throughout sub-Saharan Africa.
In some areas of Western Kenya access to inputs is hampered by lack of preferred inputs, late
1
Rural population birth and growth rates have eased of late, in part due to better education and increased
burdens on civilians to pay for health and education services.
3
delivery, and high costs of inputs. Marketing constraints are given high priority by farmers in W.
Kenya and are visible on the landscape through the absence of higher value crops.
Profitable agricultural opportunities are not a sufficient condition for good land management on
farms. The prevention of degradation, in the absence of traditional techniques of fallowing, requires
new innovations and the sharing of information. On the technical side, soil fertility replenishment,
mitigation of land degradation, and enhancing soil organic matter must be accompanied by
appropriate conservation practices, crop diversification and increased planting of trees on farms: in
short, good land husbandry. An integrated approach to management of the agricultural landscape is
required. More sustainable agriculture will in turn provide environmental benefits that accrue at the
local, national, and global levels. Improved agricultural practices must also increase farm
profitability, which is essential if they are to be adopted by farmers. Recognition of the social and
economic needs and expectations of rural populations must be an integral part of any proposed
changes in agricultural practices. On the policy side, focus has been on the larger farmer and ensuring
adequate food supplies to urban areas (e.g. packages centred around expensive seed and fertilizers).
Similarly, flows of information are generally poor in smallholder rural communities. Flows from
research and extension to communities are inadequate, as are flows between households and within
households (i.e. from men to women). In W. Kenya this is not so much due to a lack of information
providers, because there are numerous NGOs complementing the extension service, but a lack of
overall planning and coordination.
CURRENT PROBLEMS IN WESTERN KENYA
At the farm household level, trends of declining agricultural productivity and declining environmental
quality have led to the emergence of poverty and pessimism towards agriculture, which now become
problems themselves in a dynamic sense by reducing the number of feasible options for improving
livelihoods. Many households have since disintegrated socially through individual migration and
diversification of livelihood strategies. Consequently, agriculture tends to become more marginalized
and efforts to invigorate productivity and reverse degradation have to be intensified. The government
of Kenya has a draft poverty eradication plan, but it is very poorly funded.
Even if the economic climate for agriculture is improved, certain types of degradation may still occur
because they take place or originate on land that is not farmed (e.g. abandoned land, roadsides, river
banks). Such situations require collective action to solve, whether that be among households within a
village or among different villages. The hilly and sloping topography of W. Kenya contributes to
trans-farm degradation. The centrally controlled style of governance in Kenya generally hinders
communities from taking their own initiatives as authority for initiative is vested in few office
holders. Recent efforts by programmes in the Lake Victoria basin offer new platforms for bringing
communities together, but these are still nascent.
KARI and ICRAF have been working on these problems in western Kenya for the past 10 years.
Several agroforestry practices exist that have been proven to be helpful with overcoming soil fertility,
weed, and erosion problems, particularly when these practices are combined with other conservation
measures (e.g. minimum tillage, integrated pest management, soil fertility recapitalization).
Agroforestry provides reasonable options for small-scale farmers to re-establish the productivity of
their land, diversify production, and reverse the downward spiral of poverty and environmental
degradation. The “Pilot Project on Soil Fertility Replenishment and Recapitalization” initiated in
1997, has begun the work of scaling up the results of research through community-led activities in
partnership with the Ministry of Agriculture and Rural Development (MoARD), local and
international NGO’s, and community-based farmer organizations.
The GoK has funded this project since its inception at a rate of $250K per year. In this project, 17
pilot villages with 2035 households in two administrative districts are participating, Siaya and Vihiga.
In these villages, adoption rates of agroforestry technologies for soil fertility improvement, including
4
improved fallows and biomass transfer of Tithonia diversifolia (a green manure system) are on the
order of 60-70%. Through a collaborative network of partners, another 10,000 farmers scattered in 16
other districts in western Kenya, have been reached and impacted. Farmers are now adding value to
improved soil fertility by growing high value crops (vegetables, fruit trees), and those who can afford
it are beginning to raise dairy animals. These technologies have certainly had profound impacts on
rural food security, incomes, and their general welfare, and this is currently being monitored to
quantify the nature and magnitude of these impacts. Constraints to adoption have been lack of
information and awareness about technology, adequate supply of seeds and planting materials,
training and follow- up. These are problems that will be addressed over the course of this project.
In addition to solving these local problems of poverty and natural resource degradation, better farming
practices including agroforestry also provide global environmental benefits. The recent Land-Use,
Land-Use Change, and Forestry Report (2000) of the Intergovernmental Panel on Climate Change
(IPCC) has identified conversion of degraded croplands into agroforestry as the land-use practice in
the tropics with the largest potential to sequester C. Estimates of C accumulation rates range from 2
to 9 tonnes per hectare per year, depending on the climate and the nature of the agroforestry practice.
Agroforestry in western Kenya can generate important global benefits in the area of international
waters by decreasing the impacts of poor land management practices on water quality in Lake
Victoria. The area that is proposed for this project in western Kenya is part of the Lake Victoria
basin, whose products and services support some 25 million people in Kenya, Tanzania and Uganda.
Impacts to water quality in this lake also have major downstream effects in The Sudan and Egypt.
Environmental degradation in the uplands inevitably affects the lake, resulting in declining fisheries
and increased infestation by the exotic aquatic weed, water hyacinth (Eichhornia crasipes). For
example, upland erosion in the Nyando River watershed in Western Kenya causes a sediment plume
in the lake that is visible from space. The other rivers (e.g., Yala, Sondu) discharging into the lake
show similar effects of inappropriate land management practices in the watersheds. KARI, ICRAF
and partners have been involved in the “Improved Land Management in the Lake Victoria Basin
Project”, which concentrates on the Nyando and Sondu-Miriu river basins that empty into Winam
Gulf of Lake Victoria. This project seeks to decrease the significant sediment loads delivered to
Winam Gulf through improved land management practices, restoration of riparian vegetation, and
restoration of the filter function of wetlands.
Agroforestry can also enhance biodiversity and agrobiodiversity in the agricultural landscape. Studies
conducted by the Alternatives to Slash and Burn Programme in the humid tropical areas of Africa,
Southeast Asia and Latin America show increased diversity of flora and fauna with the adoption of
agroforestry practices. Increased heterogeneity on the landscape creates more niches and increases
habitat for different species. Agroforestry also has the potential to contribute to biodiversity in
protected areas by providing wood to rural households and thus decreasing pressure on resources
inside preserves. It is recognized that this benefit is context specific, but there are situations in the
proposed project area where this may apply. Finally, agroforestry affects belowground biodiversity
(agrobiodiversity) in ways are only beginning to be understood. For example shifts in nematode
populations in improved fallow systems and communities appear to be more diverse and more even
(Desaeger et al., 1999). This increased evenness appears to decrease the pathogenicity of nematodes
on subsequent crops. Other areas of belowground biodiversity still need to be explored.
3. LINKAGE S TO CAS AND NPEP AND BANK PROGRAMS
The World Bank does not yet have a finalized Country Assistance Strategy (CAS) for Kenya. The
draft CAS (1998) notes that poverty levels are increasing rapidly in Kenya and poverty levels in rural
areas are 46%. The Strategy states that increasing poverty and the widening gap between rich and
poor in Kenya pose the greatest threat to political stability. The Strategy matrix specifically identifies
actions to decrease poverty, which includes improvement of agricultural service delivery to farmers.
The proposed project is consistent with the draft CAS strategy, particularly with the focus on
5
community-based initiatives in the fight against poverty. The draft of the National Poverty
Eradication Plan (NPEP, 1998) places emphasis on the high and medium potential areas of Kenya,
which characterize the Lake Victoria Basin. Priority is given to these areas because of their high
population density, high incidence of rural poverty, and the stagnant economic growth. Among the
interventions specifically targeted in paragraphs 7.10 and 7.11 are soil conservation and agroforestry
as means for raising productivity, diversifying production, and raising farmers’ incomes.
4. ONGOING AND PLANNED PROJECTS IN THE REGION
GOK supports a number of programs that assist farmers in poor, rural areas of the country. While all
of these programs incorporate the concepts of financial, social and environmental sustainability, they
are generally applied at the on-farm or community level. Thus, while these programs provide
assistance for activities related to GEF Operational Programs on integrated ecosystem management,
international water bodies and forest biodiversity, many of them also provide the basis for effective
outreach mechanisms for working with smallholder producers and a broad range of stakeholders,
including government, private sector and NGOs. All such programs utilize participatory methods and
community-based management approaches. Thus, this GEF project will build on the advances of these
other projects that are operating in the region, use the networks that have been created to reach
farmers, and will work with these projects to advance the IEM approach.
A consortium of agricultural research and development partners for western Kenya was launched in
January 2001. This consortium comprise 40 organizations that are active in research and development
in western Kenya among them organizations/agencies such as KARI and KEFRI, the extension
branch of the MoARD, NGOs, CBOs, and farmer groups and associations. This Consortium of
research and development partners will be involved in designing the project to facilitate achievement
of the goals of the project and to ensure the involvement of key stakeholders. To backstop the
Consortium, the project will strengthen and continue to provide the following services: a) training
development agents in participatory methods and technical aspects of improved agriculture; b)
creating awareness through field days, visits and tours for the partners; c) participation in annual
agricultural shows in various locations; d) production of extension and training materials for partners
and farmers; e) establishment of tree seedling nurseries, f) organization of regular joint planning
meetings; and, g) support jointly with GTZ the publication of a quarterly newsletter – “Miti ni
Maendeleo” meaning “Trees for development” - Integration of Trees into the Farming Systems.
Rural development programs are coordinated through the MoARD, as well as with other ministries,
provincial and municipal institutions, NGOs, and farmer organizations. A list of the principal rural
development programs that explicitly incorporate natural resources management concerns and that are
relevant to the design of this project is presented below. Under the proposed Block B Grant, this list
of programs will be up-dated and their activities reviewed to insure that this project complements
these activities.

Agricultural Technology and Information Response Initiative (ATIRI): This is a national
project targeted at increasing the agricultural knowledge base of smallholder farmers in the
country by facilitating the dissemination and adoption of agricultural technologies. KARI
manages the initiative on behalf of a National Steering Committee (NSC), composed of
representatives from MOARD, non-governmental organizations and other stakeholders. ATIRI
was initiated in June 2000 and is expected to continue until June 2002 with a possible extension if
further funding will be forthcoming. It is funded by the International Development Agency
(World Bank) and is designed to contribute to the achievement of the goals of the 2 nd National
Agricultural Research Program (NARP II), which are to ensure food security, reduce poverty and
move towards sustainable natural resource management. The objectives of ATIRI are to improve
farmers’ ability to make demands on agricultural service providers and to enhance the
effectiveness of intermediary organizations and farmers groups in meeting the knowledge needs
of their clients and members. Twenty-three out of 80 funded proposals come from Western
6
Province from the mandate area of two KARI centers involved in the pilot phase, Kakamega
Regional Research Center (RRC) and Kisii RRC. Proposals range from technology transfer so as
to increase production of food crops such as groundnut, soybean, maize, bean, rice, sorghum,
banana and vegetables using improved seeds and fertilizers (inorganic and organic) to improved
dairy and small livestock farming (ATIRI, 2000).

The Lake Victoria Environmental Management Project (LVEMP) is a tripartite project aimed
at rehabilitating the ecosystem of Lake Victoria for the benefit of riparian communities, the
national economies of which they are a part and the global community. The project was started in
1997 sand will be completed in 2003 with joint funding from GEF, IDA and the three East
African governments (Kenya, Uganda and Tanzania). It is implemented by three national
secretariats each headed by a high level officer selected by the respective governments. The
specific objectives of LVEMP are to maximize sustainable benefits to riparian communities using
resources from the lake, conserve biodiversity and genetic resources for local and global benefit
and to harmonize the national management programs of the three riparian countries. Achievement
of these objectives would help reverse the increasing threat of environmental degradation to the
lake and its adjoining catchment area (184,000 km2). Specific project activities range from
integrated development of groundwater resources and reduction of sediment and nutrient flow to
management of the water hyacinth infestation to managing lake pollution and water quality
including strengthening and harmonizing national regulatory and incentive frameworks. The
project has invested heavily in capacity building (100 regional MSc, 15 Ph.D. and more than 2000
short-term courses) and collaborates widely with Universities, Government agencies and local
communities in the three countries to implement the project (LVEMP, 1996).

Soil Management Project (SMP). This national project was started in 1994 with co-funding
from the Rockefeller Foundation and KARI. The main objectives of the project were to appraise
causes of declining soil fertility in smallholder farms and to develop low cost technologies for
addressing the problem together with farmers. The SMP was the forerunner of the Legume
Research Network Project described below. It adopted a farmer participatory research approach to
ensure that farmers and other stakeholder participated fully in technology development and
transfer. In addition to decline in soil fertility, farmers identified a lack of suitable crop varieties
and livestock feed as other important constraints affecting farm productivity. The second phase of
the projects commenced in January 2001 and it is to take five years. The main activities to be
undertaken include: a) scaling up of promising technologies developed in Phase 1, and b) research
on integrated soil management to address gaps identified in Phase 1 and soil fertility constraints in
new sites. A third objective c) is to introduce and evaluate crop varieties using the methods of
participatory breeding (SMP and LRNP, 2000).

The Legume Research Network Project (LRNP) is a national project with a goal to promote
the use of green manure/cover crop legumes so as to improve the productivity of smallholder
farms and to conserve the environment. The broad objectives of the project were to identify
legumes that are adapted to specific climatic conditions, to conduct on farm research so as to
integrate the legumes into their farming systems and to bulk the seeds of the most promising
legumes so as to alleviate seed shortage(s). Network members were and still are drawn from
KARI, NGOs and the University of Nairobi. The collaborators are staff from the MOARD and
farmers in different regions of the country. The project’s research sites span the country from sea
level, to the highlands west of the Rift Valley in Nyanza (Kisii) and Western Province
(Kakamega). At the end of the first phase of the project (December 2000), 40 legume species had
been screened at 11 sites with different climate characteristics and the effect of inoculation with
Rhizobium spp. on the performance of selected legumes had been evaluated. The second 5-year
phase of the project (2001-05) will continue research on integrated nutrient management and
green manure legume residue management. Nitrogen release patterns, potential of various green
manure legumes to fix nitrogen, identification of spatial and temporal niches for introduction of
green manure legume on-farm and utilization of green manure legumes as livestock feed and for
7
human consumption will also be tested. Other activities will include dissemination of green
manure technologies and seed bulking, storage and exchange (LRNP, 1999).

DFID Rural Credit Project. DFID is putting into place a subproject that will address the
constraints to fertilizer purchase and test various options for rural credit. This project will support
acquisition and testing of promising new crop varieties that will take advantage of improved soil
fertility. The subproject will be coordinated through the Pilot Project and has begun in March
2001.

Other Projects. The Kenya Agricultural Research Institute is also undertaking numerous small
studies in Western Kenya in the mandate areas of Kisii RRC and Kakamega RRC. It is not
possible to describe the projects in detail here, but suffice it to say that for Kisii RRC, they
involve improvement (breeding) and production of sorghum, maize, groundnut, cotton, rain-fed
rice, cassava and breeding of livestock. In Kakamega RRC, they involve improvement (breeding)
and production of sorghum, millet and rice. In addition, a comprehensive Geographical
Information Systems (GIS) database has been established for fertiliser recommendation domains
in Kisii using climate data from KARI-CIMMYT’s Maize Data Base Project, soil data from the
Kenya Soil Survey (KARI) and fertiliser trial results from KARI’s Fertiliser Use
Recommendation Project (Wamae et al., 1999).
. The following full-sized GEF projects (primarily in the biodiversity and international waters focal
area) are currently being executed in:

WB Tana River Primate National Reserve Project;

WB Lake Victoria Environmental Management Project (LVEMP);

East African Cross-Border Biodiversity Project.
5. PROPOSED PROJECT
A) Objectives
Kenya has just completed preparing its National Poverty Eradication Plan for the period 2001-2015.
Smallholder agriculture will be a key area targeted, as rural poverty rates are extremely high (30-50%
of the households in the districts of the two targeted provinces). Reversing land degradation and
attaining food security will be the first steps in overcoming rural poverty. In addition, western Kenya
has a relatively high incidence of AIDS, which has had and will continue to have serious impacts on
the rural labor force, further worsening these problems. The principal national objective is to
integrate poverty reduction activities directed at the smallholder farmers within the context of an IEM
approach which is consistent with addressing Kenya’s priorities for degraded land rehabilitation,
adaptation to climate change, biodiversity conservation, and conservation of Lake Victoria. These
objectives are in line with the GEF objectives under OP 12.
The key objectives of the project are to reverse land degradation and promote income generating
activities for rural farmers in western Kenya, increase C sequestration in soils and vegetation in the
agricultural landscapes of western Kenya, contribute to reducing pollutant loads into an degrading
international water body (Lake Victoria), and reduce the loss of biodiversity. Specific objectives are:

Enhance the potential to sequester additional above- and below-ground carbon (including
total greenhouse gas (GHG) accounting), and develop scientifically sound and cost
effective procedures and protocols to measure, monitor and validate above- and belowground sinks in different land-use systems in accordance with CDM guidelines;
8

Using a Community Driven Development approach in the project area, scale up improved
management of soil fertility and agroforestry activities, promote activities to rehabilitate
degraded lands, and introduce value-added cropping systems to promote food security,
poverty reduction and the permanence of above- and below-ground C in agroecosystems
simultaneously;

Reduce the negative impacts of agricultural activities on Lake Victoria by improved land
management and agroforestry, improved erosion and sediment control, and decreased
nutrient delivery to watercourses that feed the Lake;

Improve the on-farm and off-farm biodiversity through agrobiodiversity and decreased
pressure on “natural habitats” (remnant forests, riparian areas, wetlands, etc.) through
improved land management systems;

Improve capacity for local communities, national, and international institutions to
implement projects that combine local and global benefits;

Examine the synergy between sustainable agricultural development and global
environmental benefits such as mitigation of GHG accumulation in the atmosphere,
forestry, biodiversity loss, and degradation of international waters;
These objectives will be achieved through the adoption of an Integrated Ecosystem (Natural
Resources) Management (IEM) framework in policy, planning and implementation of development
activities in selected districts in Western and Nyanza Provinces and part of Rift Valley Province. In
meeting these global objectives, the project will increase sustainability of farming systems and
thereby improve the livelihoods of rural farmers. The assessment of different land use systems for
their contribution to soil carbon sequestration and the implications of this for gaining carbon credits
has already attracted the interest of some Swiss multinational companies, who would like to gain
experience on working with rural communities on the one hand, and international agencies like the
GEF and the World Bank on the other, to assess the potential benefits of the Kyoto protocol.
Expected outcomes of the GEF Project are:

An enabling environment to facilitate the adoption of IEM approaches in national,
provincial, and local development planning;

Strengthened institutions and increased institutional capacity to identify opportunities,
formulate, and implement policies in support of integrated ecosystem management
approaches;

A range of investments identified, coordinated, and implemented through the
development of IEM production plans.
These outcomes will be verified through the measurement or description of impact indicators to be
developed during the project design phase covered by the PDF Block B grant. During this period (12
months), indicators will be developed to measure the following trends and/or achievements:

Improved livelihoods of small-scale farmers;

Increased community involvement in IEM (assessment, planning, decision making,
implementation, evaluation, etc.);

Increased soil C in cultivated and grazed areas;

Increased aboveground C (plant biomass) in cultivated and grazed areas;
9

Reduced sediment and nutrient loads in watercourses draining into Lake Victoria;

Increased incidence of endemic species in agricultural landscapes;

Reduced pressure on protected areas and other biodiversity-rich areas of global
importance;

Results and lessons learned disseminated internationally.
B. Project Description:
An Integrated Ecosystem (Natural Resource) Management (IEM) Approach
The proposed project is consistent with the GEF OP-12 because it supports the reduction of net
emissions of greenhouse gases and increased storage of carbon in terrestrial ecosystems, the
conservation and sustainable use of watersheds, protection of an important international water body,
and the conservation and sustainable use of biological diversity through an IEM framework. The
proposed project would support interventions that specifically address the following constraints that
impede the adoption of IEM approaches in Kenya:

Absence of necessary data and information required by resource managers, planners and
decision-makers to mainstream an IEM-based approach into production activities;

Weak policy framework and enabling environment supporting the adoption of IEM
approaches;

Weak institutions at national, regional, and local levels with weak capacity to adopt and
implement policies formulated in support of IEM objectives;

Insufficient technical assistance and financial resources to reduce the perceived risks
faced by resource managers in the decisions leading to the adoption of non-traditional
land management strategies in support of IEM objectives;

Difficulty in integrating activities related to sustainable ecosystem management that
transcend municipal and provincial boundaries because of lack of co-ordinated planning
across these boundaries.
The overall implementation strategy of this project is geared to help in the development and adoption
of an (IEM) approach to the problems of rural poverty, environmental degradation, and the increased
risks and uncertainties that derive from changes in the biophysical and socio-economic environment
(Figure 1). Society relies on ecosystems not only for the continued production of “goods” or
commodities, but also for the maintenance of critical “services” that ecosystems provide. Where goals
for production of ecosystem goods and services often conflict with one another, IEM is a means of
achieving the goals of poverty alleviation, increased food security and environmental protection. The
overall goal for this approach is to improve agroecosystem performance in terms of biological
productivity, integrity, maintenance and perpetuation (doing so over the relevant spatial and temporal
scales) while at the same time ensuring that these improvements can be adopted by farmers and
decision-makers at various levels and that they actually result in poverty alleviation and farmers
empowerment. A multi-scale, multi-objective, systems approach is needed to fulfil such an objective.
This approach also needs to embrace the competing interests of the various stakeholders, and key
interconnections and thresholds for ecosystem services. At its core lies a search for a balance between
productivity of agriculture, forestry and fisheries and resilience of these systems (or growth and
adaptability), between increases in the adaptive capacity of agroecosystems (including that of their
10
Figure 1. Integrated Natural Resources Management Framework
managers at different levels) and increases in the productivity of these systems. The output of such an
approach are ranges of flexible and adaptive options, for farmers and other resource users, as well as
for policy-makers at various levels. The IEM framework is based on the premise that there are social,
economic, and biophysical interactions between the goals for production of environmental goods and
services that are desired by different stakeholders. Reconciling conflicting goals and uses of land is a
critical challenge for land management. Understanding how land-use decisions and management
practices affect the production of different ecosystem goods and services is necessary for sustainable
management of the agricultural landscape.
11
This framework includes both biophysical, socio-economic and policy analyses and interventions in a
comprehensive approach to solving problems of rural poverty and providing goods and services to
different stakeholders. Problem analysis involves understanding both biophysical and socio-economic
components of poverty and land degradation as well as the temporal processes and the cause-effect
chains that result in the current situation. Options for enhanced biological productivity, enhanced
human well-being, and enhanced ecosystem integrity and resilience are developed in parallel and
tradeoffs are analyzed. The information on causes and effects, potential options for addressing
problems at the farm and landscape scale and tradeoffs between different options and production
levels are then brought to negotiations between different stakeholders and used to inform the
negotiation process. Developing this framework and making it operational will empower the
communities in the project area to take on the long-term responsibility for managing their own
resources. Representatives from farmer groups have indicated that they would prefer communal rather
than private benefits from the project e.g. improved water systems, schools, roads, rather than direct
payments. This may require that something like a “village trust fund” be established.
Implementation arrangements
The project will build on activities begun by KARI, ICRAF, and KEFRI in western Kenya and will be
carried out using a using the principles of participatory research and development that are consistent
with the model for IEM (Izac, 2000). Working with small-scale farmers through community driven
initiatives will demonstrate the feasibility of undertaking projects with clear development objectives
that also fulfil objectives of international conventions (e.g. CBD, FCCC, CDD). This project will also
contribute to GEF objectives associated with international waters.
KARI, ICRAF and KEFRI are currently working in western Kenya to promote sustainable farming
practices and to bring scientific innovations to thousands of small-scale farmers. Responsibility for
implementation of this project will be shared jointly between KARI and ICRAF. Consultations with
regional and national leaders in the Ministry of Agriculture, KEFRI, local NGOs, and farmers groups
in Nairobi at a stakeholder’s workshop at ICRAF headquarters (December 7-9, 2000) resulted in
endorsement of the project from these key stakeholders.
C) Project Intervention Zone
Geographic Foci
12
This project will focus on the lands in western Kenya that fall within the Lake Victoria watershed.
Figure 2 shows the limits of the watershed, the provinces in the watershed, and has a second layer
overlaid to show the high rainfall areas within the Kenyan portion of the Lake Victoria watershed.
The proposed GEF-supported project would finance the incremental costs to support an IEM approach
in 35-40 divisions. Project activities would be carried out in bounded production ecosystems to be
confirmed during project preparation. A list of candidate districts will be drawn up taking into
account (i) general guidance through the eligibility criteria of GEF under OP 12; (ii) potential to
sequester C (using the carbon deficit idea of Walsh et al., in press); (iii) relationship between land
degradation and water quality degradation in Lake Victoria; (iv) potential to increase biodiversity in
agricultural landscapes; and (v) the proximity to reserves with significant degradation due to external
pressure. The methodology for prioritizing districts and a set of preliminary criteria for
site/ecosystem are presented in Annex 1. Criteria for selection of specific production ecosystems and
predominant landscapes will be developed further and applied during the initial phase of project
preparation.
Proposed Components
A preliminary description of components based on the current assessment of the ongoing activities in
the region and the required activities to achieve the proposed GEF objectives is described below.
Project components will be defined more precisely during the participatory PDF preparation process.
All planning, policy, research, and monitoring activities of the project will be developed to support
field activities in the selected project sites.
Component 1. Development of decentralized plans for IEM
GEF resources for this component will be used to fund incremental costs of studies, community
Western Province
Nyanza Province
Rift Valley Province
Precip > 1300 mm
Lake basin boundary
Figure 2. Lake Victoria basin, provinces and rainfall regime
in western Kenya.
13
meetings, technical assistance, information campaigns, and stakeholder workshops to formulate IEM
action plans. Theses funds will be used to complement the ATIRI, LRNP and SMP-funded activities
to develop and implement a Participation Action Plan (PAP). A PAP would strengthen the integration
of stakeholders including smallholders, producers, NGOs, local governments and others whose
activities impact significantly on natural resource management by promoting their participation in the
policy and decision-making councils at the local, provincial, and regional levels. This activity would
comprise the basis for participatory planning and collective implementation of the IEM framework
that would identify critical problem domains and propose solutions through collaborative land and
water management and conservation activities. The outputs from this component would be technical
targets, and would be supported by individual rural community land-use plans (which would be
presented by beneficiary groups) and by plans for collective activities (e.g. at the first-order watershed
scale). Rural extension, private and public sector providers of technical assistance would also be a
pivotal element of the proposed alternatives as it aims to communicate with all participating farmers
and communities a technical strategy for implementing sustainable agricultural projects generally at
the farm level.
Component 2. Mainstreaming and scaling up IEM interventions
GEF funds will fund incremental costs for extension and participatory research to promote wider scale
adoption of sustainable farming practices and then move farmers to value-added land-use systems.
This approach would thus promote food security, poverty reduction and the permanence of local and
global environmental benefits simultaneously. Resources would be used to fund participatory
adaptive research with farmers, farmer-to-farmer exchanges, training of extension workers and rural
development practitioners (NGO’s, MOARD Extension Branch, local development authorities, etc.).
Extension messages will be developed
Component 3. Sustainable farming investments
GEF funds will finance incremental costs of technical assistance, goods, and in some cases, small
projects identified in the IEM plans. These activities will complement the financial and technical
assistance for economic infrastructure and diversification of production provided under ATIRI and
SMP. Small project preparation would be assisted by extension agents and reviewed by district units
that would have the necessary technical capacity to evaluate subprojects. The standard small projects
will establish a baseline for identifying activities eligible for GEF financing, including potential
starter activities. Possible activities which could be supported under this component include
development of village nurseries to provide trees for agroforestry interventions, development of
existing agrobiodiversity resources, support for economic activities that could serve as alternatives to
existing practices which threaten important landscape elements such as in protected area buffer zones
and ecological corridors or where current practices are highly destructive and extensive, etc.
The GEF-supported activities would be characterized by: (i) demand-driven small project selection
based on proposals prepared and submitted by eligible project participants; and (ii) decentralized
project review to facilitate rapid response to assistance requests and maintain closer contact with
potential participant needs.
Component 4. Monitoring and evaluation
GEF funds will be used to support incremental costs of developing a detailed monitoring and
evaluation (M&E) system, particularly with respect to global environmental services of C
sequestration, international waters conservation and biodiversity conservation. M&E will be carried
out using participatory mechanisms, however a much stronger technical/scientific component
associated with global benefits would be integrated into both the measurement and evaluation of
results, building on methods generated in the targeted research component of the project. [Note:
14
Because the combination of DRS and remote sensing will be an important part of baseline
development and of the M&E system, details of these techniques will be presented in a technical
Annex, Annex 2]
Component 5. Project administration
Project administration will be built on the foundation of the consortium for agricultural R & D in
western Kenya. The consortium has a coordination committee of 7 institutions, including the local
councils, local representatives of the Ministry of Health's HIV/AIDS program and the Forest
Department. This arrangement will allow the project to operate in a larger number of locations and
effectively cover the 20 districts and the 6.0 million smallholder farms in western Kenya, many
farming in soils of low fertility.
The implementation structure of the project will have three components: a) an inner small core
committee of institutions overseeing the day-to-day coordination of the project, b) an outer larger
group of institutions that make up the consortium for wide-scale dissemination and, c) national
steering committee of institutions that will guide and oversee the overall development of the project.
The composition of the respective committees will be as follows: a) inner core coordination
committee (7 institutions) – KARI (Kakamega), KARI (Kisii), KEFRI, ICRAF, MoARD (Nyanza
Province), MoARD (Western Province), CARE, Vi Agroforestry Project, local NGO (Ugunja
Community Resource Centre) and MICWP (Maseno Inter-Christian Child Welfare Project – a local
NGO dealing with HIV/AIDS affected households). This coordination committee (secretariat based
at KEFRI station at Maseno) will co-opt other members form the larger consortium if and when
necessary. The national steering committee will include the following institutions: KARI, KEFRI,
ICRAF, Ministry of Agriculture and Rural Development (MoARD – the Director of Agriculture or his
representative), National Environment Secretariat and Office of the President representative. The
presence of the latter is essential for mobilization support from the district and provincial
administration.
MoARD would be expected to play a key role at the regional and sub-regional level (several districts).
This is because the ministry has an appropriate mechanism to reach out through the newly instituted
National Agricultural and Livestock Programme, funded by SIDA. This operates in focal areas
(catchments) in districts for one year and then moves on to another location, leaving behind skeletal
technical staff for backstopping and monitoring impact. This is good process that integrates the
various institutions (NGOs, CBOs,etc) in that area, thus enhancing impact. The sub-regions will
have their own committees (made up of partners existing in that area) in which the District
Agricultural and Livestock Development Office could be the conveners.
6. EXPECTED LOCAL AND NATIONAL BENEFITS
.
Local significance
The highlands of western Kenya represent 15% of the total land area of the country and accommodate
40% of the country’s population. High population density and unsustainable land-use practices have
led to degraded agroecosystems and declining productivity. Agriculture is not meeting the needs of
the people. Moreover, with a population growing at the rate of 3.4% per annum, the situation with
respect to food security and environmental degradation is likely to become even more severe without
outside intervention.
This project will contribute significantly to mitigating these problems and to improving the
livelihoods of the people in western Kenya. Since fertilizers cost 2 to 4 times more at the farm gate in
Africa than they do in Europe or North America, different approaches are being developed that
15
combine organic and inorganic inputs. A promising one is a robust NRM approach which brings
nitrogen and carbon from the air and phosphorus from indigenous phosphate deposits, together with
biomass transfers of nutrient-accumulating hedge species. Leguminous tree fallows of several species
of Sesbania, Tephrosia, Crotalaria, Cajanus, and others accumulate up to 100 kg N/ha in one year,
which is incorporated into the soil before maize planting.
These fallows also provide multiple benefits such as on-farm fuelwood production, capture of leached
nitrate, recycling of other nutrients, Striga control, better soil physical properties, reduced erosion and
carbon sequestration. Tree fallows provide the basal application of N, which can be supplemented by
top-dressing applications with N fertilizers if conditions warrant. Experiments in western Kenya
show that a single basal application of P or smaller annual application of Minjingu phosphate rock (a
local deposit) can triple maize yields and appears to be 70-80% as effective as imported triple
superphosphate in many cases. Tithonia diversifolia has high nutrient concentration in its leaves
(3%N, 0.3%P, 3%K) and decomposes rapidly in the soil. Biomass transfers of Tithonia at rates of 2 5 tonnes of dry matter per hectare routinely double maize yields without any fertilizer additions.
Because of these technologies, thousands of farm families are becoming food secure and no longer
suffer from hunger periods as maize grain yields increase. Soil fertility replenishment is a multidisciplinary issue involving social, economic, and policy studies as well as the biophysical dimension.
The elimination of this root cause of food insecurity in Africa augurs well for turning around the
downward poverty-environmental spiral on this continent.
In addition to productivity benefits of mitigating soil degradation through agroforestry, there are many
options for diversification of production through domestication of medicinal trees or local fruit trees.
Work in Shinyanga, Tanzania, has identified more than 300 trees species used for the treatment of
more than 100 human diseases; some of these species are also indigenous to western Kenya.
Additionally, Tithonia is used in the preparation of a number of traditional medicines. These options
may be helpful in decreasing risk for farmers and increasing income generating options.
There is an additional opportunity that will be examined during the planning phase of this project. The
University of Maseno is working closely with local traditional healers and birth attendants to collect
traditionally used medicinal plants and validate the safety and effectiveness of formulations. This
project offers a unique opportunity to include the documentation of traditional healthcare treatments,
and identify opportunities for the cultivation, processing and marketing of affordable phytomedicines
by local communities. KARI, ICRAF and other partners have an opportunity to work closely with
healers and farmers to identify sustainable cultivation practices as part of the proposed agricultural
intensification..
Expected outcomes of this project that are relevant to national objectives are:

Better farmer-led land management practices that increase productivity while contributing
to global environmental benefits;

Enhanced food security and livelihood systems for the local rural, peri-urban, and urban
communities;

Rural development strategies at national, provincial and local levels which explicitly
integrate ecosystem concerns – including targeting, prioritization of activities and ex-ante
and ex-post evaluation criteria;

Extensionists and producers will learn to explicitly factor ecosystem impacts into their
evaluation of technical options;

More effective prevention and control of land degradation;
16

Local social organizational structures (village and rural community level) which are able
to evaluate and address ecosystem concerns;

Replicable, locally adapted production systems and technologies which support both
poverty reduction and improved ecosystem management objectives;

Results and lessons learned disseminated in the country
7. EXPECTED GLOBAL BENEFITS
Global significance - Climate change
Kenya ratified the United Nations Framework Convention on Climate Change (UNFCCC) in 1994.
UNFCC aims to achieve the stabilization of the concentrations of greenhouse gases in the atmosphere
to acceptable levels. Kenya recognizes the relevance of land use and restoration activities within its
national boundaries as major factors, which could contribute to achieving UNFCCC goals.
Assessment of agricultural activities in the light of their carbon sequestration functions indicates that a
significant portion of these activities will be carried out in Kenya within the context of smallholder
agriculture and in the high potential areas of western Kenya.
The replenishment of nitrogen and phosphorus may have important effects on changes in carbon
stocks. The loss of topsoil organic carbon associated with soil nutrient depletion has been estimated
at an average rate of 0.22 tonnes of C per hectare per year. Yields on typical degraded sites in western
Kenya range from 0.5 to 2 tonnes of grain per hectare per year. When soil fertility is replenished
maize grain yields increase around 4 to 6 tonnes per hectare per year and C sequestration rates
become positive, averaging as much as 1.5 tonnes of C per hectare per year. When more trees are
planted on field boundaries, in fields, and as orchards C sequestration rates increase further. Based on
this work the recent IPCC special report Land Use, Land Use Change and Forestry (IPCC, 2000)
indicates that C sequestration rates may range from 1.2 - 5.1 tonnes of C per hectare per year, with a
modal value of 3.1, the highest value reported for any agricultural practice by that report.
Nutrient-depleted fields have little biomass C stock; a time-averaged modal figure is in the order of 23
tonnes of C per hectare, virtually all below ground. Soil fertility replenishment practices based on
improved fallows, rock phosphate, biomass transfer and soil conservation for 25 years are
conservatively estimated to result in time-averaged C stocks of 32 tonnes of C per hectare
(Alternatives to Slash and Burn Programme). Such stocks are virtually all in the soil, as crop and
fallow accumulation may account for only one tonne of C per hectare above ground. The magnitude
may be higher with the introduction of conservation tillage, which is now beginning to be researched
and adopted. When trees are incorporated after fertility replenishment, total time-averaged stocks
may reach 70 tonnes of C per hectare, which includes 34 tonnes of C in the above ground biomass and
36 tonnes of C below ground. Carbon sequestration may therefore be considerable with land-use
conversion to agroforestry systems that involve soil fertility replenishment, intensification, and
diversification of farming with the use of high-value domesticated crops/trees in subhumid areas of
Africa such as Western Kenya. This provides significant benefits towards mitigating greenhouse gass
effects on the global climate.
Global significance – International waters
The Lake Victoria Basin supports one of the densest and poorest rural populations in the world, with
population densities up to 1200 persons per square kilometer (Hoekstra and Corbett, 1995). The
17
Welfare Monitoring Survey that was conducted in Kenya in 1994 indicated that the incidence of
severe poverty (household expenditures of less than KShs 703, or US$16 per month) is between 30
and 50 percent in seven lake basin districts (Central Bureau of Statistics, 1998). Land degradation has
caused significant negative impacts downstream causing eutrophication of the lake, which in turn has
led to rapid colonization of the lake by water hyacinth, and decreased fish and aquatic plant diversity.
The economic impact of this has been great. For example, the operations to keep hydroelectric
generating turbines clean have cost Uganda $600,000 per year. The fishing industry, which employs
500,000 people in the riparian countries, has also been severely affected.
Recent research has shown that common property and grazing lands in the Nyando and Tondo river
basins are subject to high erosion rates, resulting in major sedimentation and eutrophication in Lake
Victoria (Chin Ong, personal communication). Replenishing soil fertility is a major element in soil
conservation and will contribute to alleviate the water hyacinth problem, protect threatened fisheries
in Africa's largest lake, and have an impact on aquatic plant and fish diversity.
Global and regional significance – Biodiversity and Agrobiodiversity
The conservation and sustainable use of biodiversity is a national priority in Kenya. Kenya was
among the early signatories of the Convention on Biological Diversity (CBD) and ratified the
convention in 1994. Kenya has actively participated in meetings of the Conference of the Parties
(COP) to CBD, and hosted the most recent meeting (COP-5) with UNEP in May 2000. In order to
demonstrate its commitment to biodiversity conservation and achieve its national priorities, the
government is implementing a series of initiatives. These include:

Completion of the National Biodiversity Strategy and its corresponding Action Plan;

Preparation of the first report to the COP in 1998 in accordance to the obligations under
the CBD to report on progress made in respect to implementations of articles 6 through 8
of the CBD;

Implementation by the national government of the GEF-supported Tana River Primate
National Reserve Project;

Implementation by the national government of two regional GEF-supported projects Lake
Victoria Environmental Management Project and East African Cross-Border Biodiversity
Project;

Kenya has designated several areas as important for conservation, including National
Parks, Reserves, Wildlife Sanctuaries, National Monuments, Biosphere Reserves, World
Heritage Sites and Ramsar sites.
The principles of the National Biodiversity Strategy (Ministry of Environment and Natural Resources,
2000) recognize that “population and poverty issues are the ultimate causes of biodiversity loss, and
can only be meaningfully addressed as national development goals.” Thus, poverty alleviation,
increased agricultural productivity, employment creation and population control are all key elements
in the National Biodiversity Strategy. For biodiversity conservation outside protected areas, the
strategy looks to promote conservation and sustainable use of biodiversity resources.
Agrobiodiversity is defined in the broad sense in the Strategy and includes plant species that are
sources of food, medicinal products, fibre, fodder, and oil. Agrobiodiversity is particularly singled
out in the Strategy and a key focal area and the promotion of farming practices that conserve
agricultural ecosystems is a key component of the strategy.
Finally, the Strategy recognizes
degradation of aquatic resources as a key element in biodiversity loss and recognizes impacts of
upstream resource use on downstream ecosystems. Thus, there is a clear linkage between
international waters and biodiversity themes in this project.
18
The biodiversity focus of this project will be the agricultural landscape and will seek to increase
functional biodiversity and so-called non-functional biodiversity. Functional biodiversity is that
biodiversity which provides goods and services to farmers. Non-functional biodiversity is that
biodiversity which is tolerated by rural communities in the landscape, but does not directly provide
products and services to farmers. On the innovative side, the effects of management on below ground
biodiversity will be investigated, and the importance of this to long-term sustainability will be
determined, particularly with respect to carbon and nutrient cycling.
Biodiversity in protected reserves in the project area is under intense pressure and strategies to
actively protect these resources are urgently needed. Some of the activities of this project should help
to reduce this pressure on these reserves, but this impact will be context specific (depending on
proximity of project interventions to protected areas). Also, the impact will be tenuous as project
activities will only partially satisfy wood demands of rural populations and may not provide sufficient
replacement for all of the products that are withdrawn from reserves.
There are numerous ecologically sensitive sites in Western Kenya, but because there are no
spectacular tourist-attracting parks in this part of Kenya, efforts to address conservation and
management of important flora and fauna have been largely inadequate. Nevertheless, this area has
unique habitats and biodiversity of local, national, and global significance. The following is a list of
these areas:

Kakamega Forest. The most notable area in this region is the Kakamega National
Reserve (IUCN Category II) covering an area of 4468 ha. This is the eastern most
locality of Guinea-Congolean equatorial rain forest in Kenya. Many of the Reserve's
plant and animal species occur nowhere else in the country; the avifauna is unique in
Kenya and includes several threatened species. Deforestation is taking place at a fast rate
and there is pressure to replant with faster-growing exotics. Illegal firewood collection,
grazing, and poaching are major threats, and the local human population is creating
significant pressure for more agricultural land.

Nandi. Also globally significant are the Nandi Forest Reserves (North and South) with a
combined area of 30,003 ha (IUCN Category VI) and were once connected with
Kakamega Forest Reserve. An endemic subspecies of greenbul occurs here. These
Reserves are considered internationally important in that they contain plants and animals
that are West or Central African in distribution, and in Kenya are limited to this area.
Logging and encroachment are major threats.

Others. Other notable ecologically sensitive sites facing serious encroachment pressure
include: Ruma National Park (12,000 ha) (IUCN Category II) (which has leopards and
buffalo), Yala Nature Reserve (which covers over 30,000 ha of wetland, minor lakes and
forests), Bunyala Forest Reserve (IUCN Category VI) (826 ha), Kaimosi Forest Nature
Reserve (19 ha) (IUCN Category IV), Lugari Forest Reserve (2163 ha) (IUCN Category
VI), Malaba Forest Reserve (719 ha) (IUCN Category VI), Maragoli Forest Reserve (470
ha) (IUCN Category VI), and West Kano Bird Sanctuary. It is notable that most of these
sites are small in area and are therefore important refugia and islands for biodiversity
conservation in an area that is heavily populated. To add to their small sizes, many of
these sites are either inadequately protected or unprotected, while others are just proposed
for protection. These factors make them especially prone to encroachment, destruction,
and species losses.
This Project will address biodiversity conservation through on-farm biodiversity conservation (i.e.,
agro-biodiversity), biodiversity enhancement off-farm in the agricultural landscape, and critical
ecosystems/habitats (e.g., wetlands, forest refugia) biodiversity conservation.
Soil fertility
replenishment is expected to enhance biodiversity in two ways. First and foremost, is by increasing
19
heterogeneity in the landscape there will be more niches for different types of species. This will lead
to increased aboveground and belowground biodiversity. Aside from niches directly created from
planting trees in agroforestry systems, there will be decreased need for women to go to adjacent
forests and range lands in search of fuelwood. Tree fallows and other agroforestry systems supply
fuelwood and construction wood that will contribute to satisfying the family's needs. Therefore, there
will be less encroachment of forests and woodlands in the landscape (riparian areas, upper slopes of
watersheds that are traditionally used for grazing, and national forests are currently under great
pressure). This project will lead to more woody vegetation both on and off farms and increased
biodiversity and reduced pressure on sensitive habitats. However, the more important biodiversity
objective of this project remains the maintenance and enhancement of biodiversity within the
agricultural landscape.
Synergy and Linkages to Global Conventions
This project will seek to fully exploit the synergy among the different global conventions, in
particular the CBD, UNFCCC, and the crosscutting issues of land degradation (CCD). Current
practices of agricultural expansion, fuel-wood extraction and settlement development can deplete
biodiversity and cause land degradation. Improved land management practices such as conservation
agriculture and increased planting of trees on farm help to mitigate these effects and can help ensure
sustainable gains in biodiversity conservation. Hence the project will strive to improve land
management practices, institutional arrangements, and policies and incentives in sectors of
agriculture, agroforestry and water use.
The linkage of land degradation with climate change can be seen through its impacts on biomass
reduction mainly through deforestation and loss of potential in C storage from crop cultivation and
soil degradation/erosion. Activities such as conservation tillage, agroforestry, reforestation and
enhanced forest management practices will help improve C sequestration potential and C cycling in
soils, thereby enhancing crop yields for farmers while contributing to global environmental goals.
Replicability
It is expected that the IEM approach could eventually be replicated within Kenya and in other
countries with similar agro-ecological problems. Experiences gained in farmer-led initiatives for
defining appropriate practices related to conservation and sustainable use of natural resources
(improved land and water management, soil fertility replenishment and maintenance techniques,
landscape scale planning and management) will be disseminated to local (e.g. small farmer
organizations), national and, international audiences. Dissemination to the latter will be done through
the sharing of experiences that would be conveyed by those directly involved in project development
and implementation (i.e., project personnel and stakeholders). Project support for the dissemination
of lessons learned envisioned under the proposed Component 2 would be consistent with the GEF
Outreach Strategy. Specific resources would be allocated to communicate to a wider audience the
project's objectives, activities and results. This would occur specifically through: (i) consultations and
outreach to local rural communities, farmer's associations, farmer-to-farmer contacts, extensionists,
NGOs and other stakeholders; (ii) project staff participating in outreach in country and internationally;
(iii) preparation of material for the general public; and (iv) preparation of material for media.
Sustainability
The project strategy is designed to ensure the sustainability of GEF-supported activities beyond the
GEF funding period by: (i) focusing capacity building of local technical resource services, producers
and land managers; (ii) recognizing and capitalizing on the crucial role of local governments and the
local producer and community organizations to organize, promote, monitor and assess
implementation; and (iii) utilizing existing institutional structures to organize project activities and
deliver outputs. This strategy reflects the lessons learned from projects in Kenya (including World
20
Bank financed and GEF supported projects) and in other countries that have attempted to introduce
new technical approaches in rural areas.
More specifically, the following activities and outcomes will ensure sustainability beyond the project
period: (i) the improvement of local, provincial and national institutional capacity to assess and
integrate natural resource management, carbon sequestration, international waters management, and
biodiversity values into development planning, particularly by increasing the recognition of the
importance of environmental goods and services that do not have a market value, but that may be
provided by farmers and other stakeholders; (ii) the development and dissemination of strategic
activities and investments at the farm, ecosystem and landscape levels to demonstrate that conserving
soil and water, increasing the conservation of agro-biodiversity, and diversifying farming functions
can be sustainable; (iii) increased promotion, by local communities, of activities that are
environmentally, socially and financially sustainable; (iv) the enhancement of partnerships between
rural community associations, producer associations and agricultural and environmental research and
extension services, which would prove favorable for disseminating technical knowledge and
technology, so as to increase adoption of sustainable production systems; (vi) a communications
program at the national level to inform the general public regarding the IEM approach to management
of Kenya’s land-based resources, publicize specific experiences in Kenya and explain government’s
role in promoting these activities.
Stakeholder Involvement
During the design phase of the concept note and PDF B application, relevant Kenyan institutions were
actively involved in the process. After discussions with rural development partners in western Kenya
and a stakeholder meeting in Nairobi in December 2000, the first draft of the project concept paper
was prepared. These consultations included (i) staff from Western and Nyanza provincial extension
services; (ii) the National Environmental Secretariat (the GEF Focal Point); (iii) representatives from
research and development partners active in western Kenya including Sida/NALEP, UNSO-UNDP,
GTZ, FAO, TSBF, RELMA, MICWP, SCODP, NAC; and (iv) farmers who are active in the KEFRIKARI-ICRAF Pilot Project in Vihiga and Siaya Districts.
Participatory approaches will be utilized during project preparation (PDF B) and implementation
phases, to enhance sustainability and stakeholder participation in mainstreaming IEM concepts in
smallholder development activities. A series of consultative workshops at national and provincial
levels is planned for the Block B phase, involving community-based organizations, key national and
international NGOs working in environmental and sustainable agriculture issues, and selected
municipal mayors, indigenous leaders and other stakeholders.
Building on the Pilot Project and other baseline activities to be defined during preparation, the project
would provide incremental training to all levels of society at national, provincial, and district and local
levels in participatory techniques for problem identification, priority setting, and the design of
solutions. Participatory mechanisms for planning and management of natural resources, including
access, and sustainable use, and conflict management will be developed. Other project activities are
expected to include project-planning workshops, participatory information gathering, focus groups,
interviews, and sector meetings, and would be consistent with the GEF policy on public involvement.
8. DESCRIPTION OF PDF ACTIVITIES
The design phase of the project will consider both local/regional development issues and global
environmental issues, recognising that the two sets of objectives are interrelated but sometimes at
odds, and may involve trade-offs. The challenge will be to optimise land management so that local,
national, and global concerns are addressed.
Specific PDF activities will include:
21
Baseline Studies
Baselines will be determined for the project area in western Kenya. Within the project area, specific
intervention zones will be identified for more detailed studies.
The baseline will differentiate between different categories. Each part of the baseline will consist of
two items: a) inventory of the current situation and b) anticipation of future trends.
Biophysical Baseline
Baseline information will be gathered on biophysical aspects of current land uses, assessment of the
degree of “sustainability” of these land uses, and what is the trend, using international standards such
as the FAO International Framework for Evaluating Sustainable Land Management. It will be useful
to evaluate whether current land use systems are moving towards or away from sustainability, and
identifying the cause and effect relationships. Land uses that are moving away from sustainability
will result in land degradation. Land degradation is a general term that reflects decreasing land
quality, but it can manifest itself in several ways and often more than one process is occurring in a
given field at the same time. In many cases it is essential that the process be identified, because this
assists in estimating the impact. However, in the final analyses, it is more important to note the
degree of degradation and to estimate the impacts of this on land use sustainability and global
environmental services.
Socio-economic Baseline (including institutional structure)
Baseline information on social and economic aspects of current land uses systems including land
management practices, costs, benefits, marketing, etc. will be established. Along with this, each land
use systems will be evaluated for sustainability and poverty levels. This information will be used for
targeting interventions and for evaluating progress. Poverty significantly limits farmers’ ability to
innovate, which is recognised as a constraint that must be overcome for project success.
Carbon Baseline (above- and below-ground)
Baseline information on carbon stocks in the biomass aboveground and in the soil to a depth of at
least one metre will be collected. Evaluations will also be conducted on carbon transfer among the
agroecosystems, e.g. cut and carry livestock feeding; using woodlot trees for building materials. These
data will be used to calculate the carbon balance for the project area, with and without the improved
land management technologies, such as agroforestry. The other GHGs (N2O and CH4) will not be
done as part of the PDF activities, but will be included in the project. The current and projected
carbon balance will be used for targeting interventions and for calculating the net carbon effect of the
project.
Agrobiodiversity Baseline
Comprehensive baseline information on biodiversity in the agricultural landscapes (including
medicinal plants and wildlife habitat) will be collected and evaluated for possible co-benefits with
agricultural production. This will build on the work of the IUCN KIFCON project report by Peter
Wass and Lucy Emmerton.
Baseline Data Requirements
Considerable information has already been gathered over the past decade in western Kenya, but most
of the information is scattered among and within various research organizations (primarily ICRAF,
KEFRI, KARI and Central Bureau of Statistics). The first job of the baseline study will be to collate,
summarise and analyse these data, working in collaboration with national institutions such as KARI,
22
KEFRI, the Museums of Kenya Indigenous Knowledge Unit, etc. This will identify gaps and where
necessary further data gathering activities will be required.
All information will be georeferenced where possible, so that maps showing “zones of similarity” can
be prepared as aids to targeting interventions. Data on socio-economics, land use and land use
change, land degradation, biodiversity, land tenure, and soil fertility are available to some extent.
However, very few data are available for carbon stocks and flows. It is expected that most of the work
and expense for baseline studies will have to be devoted to establish a reliable inventory for above and
below ground carbon. This information is most important for the project since it will establish the
reference point against which the impacts of the project activities on the carbon balance will be
assessed.
Project Design Based Upon Participation of Stakeholders
The project design will conform to the practices of participatory design of development initiatives that
are congruent with an IEM approach and with the draft Kenya Country Assistance Strategy.
Identification of Intervention Areas
The outputs of the data synthesis activities and baseline surveys will be the delineation of target
intervention zones and definition of target interventions to meet the project’s objectives. A
preliminary method for prioritizing districts and a set of preliminary criteria for site/ecosystem are
presented in Annex 1. Criteria for selection of specific production ecosystems and predominant
landscapes will be finalized during the PDF B phase and applied during the initial phase of project
preparation.
Design of Procedures for Scaling Up
Pilot activities have been underway in western Kenya to develop technologies suitable for sustainable
land management (for example Agroforestry) and carbon sequestration. Scaling up the work from
these pilot activities will be done in collaboration with local NGOs and farmer groups. These groups
have already indicated their interest and commitment to working on these activities and have
indicated that they are willing to accept responsibility for outreach and farmer education. Outputs
from this activity will be a detailed work plan for collaboration between KARI and ICRAF staff and
these groups, to extend these technologies to farming communities.
Design of Practical and Cost-Effective Monitoring and Evaluation Framework
A monitoring and evaluation system will be designed which can be sustained after the project ends.
Key areas will be identified for further research to support these activities during the PDF B phase of
the project. The system will be set up to monitor:

Project impacts on carbon stocks and flows, including additionality, permanence and
leakage, etc., impacts of project activities on sedimentation in several Lake Victoria
watersheds, and local biodiversity. These monitoring system will have to be based on
simple and cost effective procedures, to ensure that they can be continued after the project
ends.

Project activities in relation to schedules, deliverables, finances etc.
Design of Capacity and Awareness Building Activities
Activities during the project will include:
23

Awareness raising: Education campaigns and network facilitation will be undertaken at
local levels and sector agencies/departments directly involved in land use management in
the project area (including traditional healers’ associations).

Designing of dissemination activities: Dissemination of practical beneficial tools and
strategies for engendering sustainable land management that provides direct benefits to
land users while conserving biodiversity and critical habitats will form a crucial element
of project activities. This would provide the triggering mechanism (or the "gateway")
that will help farmers and other resource users and stakeholders to adopt sustainable land
resources management.

Capacity building: Strengths and weaknesses of sectors and agencies involved in NRM in
the area will be assessed, and rectified where needed by appropriate training activities.
Capacity building and organisational assistance are vital where transactions costs and
information scarcity limit the efficient execution of sustainable integrated ecosystem
practices by many dispersed small farmers/herders. Acting individually, landowners may
not reap maximum collateral social and environmental benefits. Hence PDF B activities
will promote organisational models that enable many household-led activities to be
managed as single, large umbrella projects. This will be done based on existing local or
project-initiated producer’s associations/organisations, or NGOs.
Design of Project Management Procedures
Management of the project will be done by the implementing organizations (KARI and ICRAF) to
ensure that project funds are managed using proper accounting procedures. Institutional arrangements
will be worked out and a management structure for the project will be designed.
Detailed Project Work Plan
This involves writing the detailed project plan, including final identification of and schedule of
deliverables, preparing a plan for reporting, etc.
Capacity Building during the PDF Phase
Capacity building will commence during the PDF B phase. An important part of the PDF B activities
will consist of a series of workshops to ensure that the project continues to have buy-in from all
concerned stakeholders. Capacity building will also include in-house capacity building of KARI
professional staff, who will conduct research and thesis work to support this project and build inhouse capacity in global environmental issues.
Consultative Workshops and Local Training Workshops
As part of the PDF B activities, two to three consultative workshops and two training workshops will
be carried out. The training workshops will be for researchers and for the more innovative, resource
farmers.
Special Studies
Examine and evaluate the rural development policy framework in Western Kenya to identify potential
impacts on project implementation and success.
Additional PDF Activities
24
Agroforestry is an important activity for promoting integrated ecosystem management. It requires
ready availability of seeds and seedlings. Seed production and tree nurseries must start during the
PDF B time frame, to ensure adequate germplasm for the project.
PDF BLOCK B OUTPUTS
The PDF activities will produce the following outputs:
Baseline Studies
The following baseline studies will be carried out:



Analysis and evaluation of data on land-use/land-cover, land degradation, existing carbon
stocks and flows, biodiversity, poverty and social well-being, sustainability of current
land use systems and interventions necessary to make these systems more sustainable
than they are today. The zones of intervention and the kinds of interventions required for
the project will be identified.
Realistic scenario will be developed for future trends on land-use/land-cover, land
degradation, biodiversity, poverty and social well-being, and carbon stocks and flows.
Realistic estimates of baseline expected carbon changes over a 30-year period will be
calculated.
Detailed Project Plan based on Community Driven Development
Intervention areas and target audience, land management treatments to be promoted including
agroforestry, soil fertility management, soil conservation etc. will be clearly identified. Procedures for
scaling-up of technologies suitable for sustainable land management and carbon sequestration will be
designed. The detailed project design will also include deliverables, time plan and schedules, and
reporting and communications procedures. Project budget, based on experience from the KARI –
KEFRI – ICRAF pilot project in Western Kenya will be included.
Monitoring and Evaluation Framework
Monitoring and Evaluation procedures for monitoring changes in carbon stock and flows, adoption
rates of improved technologies, but also impact of the technologies on the Lake Victoria Basin and
local biodiversity will be defined. Included also will be a carbon monitoring system based on simple
and cost effective means to ensure that it can be continued with local resources for an extended period
after the project.
Capacity Building and Awareness Programme
This will include a plan for training and consultative workshops.
Background study and report
The report will describe the current policy environment in Western Kenya related to policies which
may hinder project implementation.
25
A special activity
For the project to succeed there is a need to fast track production of seeds and seedlings during the
PDF phase to ensure germplasm for speedy project implementation.
9. NATIONAL LEVEL SUPPORT
The proposed project is consistent with and directly supportive of the GoK’s National Poverty
Eradication Plan, and the Rural Development Strategy (currently under preparation). The NPEP and
RDS both address the objectives of reducing poverty and ensuring food security by promoting agoindustries and exports in an environmentally sustainable manner. The GoK strategy is to promote
technologies that limit natural resource degradation and sustain environmental improvement.
Specifically, these will include programs to promote sustainable conservation and management of
natural resources mainly soil, water, forestry, fisheries, minerals, etc., through awareness creation
among communities, and the development of land use policies that will facilitate long-term
investments in farm improvement, efficient cropping and livestock production systems. The project
will also promote resource conservation through protection of water catchment areas, re-forestation,
agro-forestry, soil and water conservation. Implementation of a sound land use policy will facilitate
access to land by those who desire to invest in rural development. Supporting Kenya's efforts for
poverty reduction is the overriding objective of the Bank's country assistance strategy (CAS) for
Kenya.
Furthermore, GoK through KWS is rehabilitating and strengthening the management of fauna and
flora and protected areas nation wide and increasing the flow of benefits from these resources to local
communities and the national economy. The World Bank has supported these initiatives through the
"Protected Areas and Wildlife Services" (PAWS) Project. However, PAWS Project focused mainly
on protected areas that are major existing or potential revenue earners. Thus much of the (non-high
tourism volume attracting) critical areas in Western Kenya were not covered under PAWS. Hence,
activities under this proposal would expand on the strategies started under PAWS and particularly
strengthen local community support for biodiversity conservation through sustainable use.
JUSTIFICATION
PDF Block B activities for this project will be necessary to permit all stakeholders to participate in the
project development through the participatory IEM framework and to allow for full participation of
the stakeholders in baseline establishment.
TIMETABLE
PDF block B activities will begin in October 2001 and will be completed by November 2002.
10. PDF-B Financial Plan
The PDF-B phase of this project will be financed through what we will designate here as
direct funding sources. Direct funding involves those funds that will be directed specifically
at execution of the PDF-B activities.
Funding for PDF-B activities will be provided by GEF; ICRAF from its core funding, which
will cover staff time and operational costs; and from KARI/GoK in kind contribution
through staff time, logistical support and overhead support; and for operational costs
(associated with management and administration, monitoring and evaluation, and soil fertility
management and recapitalization trials) through the IDA Credit under NARP II.
26
BUDGET FOR PDF BLOCK B ACTIVITIES
(USD 000)
TOTAL
COSTS
GEF
BASELINE STUDIES
205.5
PROJECT PLAN
IDA
(NARP II)
ICRAF
KARI/GOK
100.0
20.0
85.0
60.0
50.0
5.0
5.0
MONITORING AND
EVALUATION
FRAMEWORK
45.0
20.0
CAPACITY
BUILDING
25.0
15.0
5.0
5.0
SPECIAL STUDY
25.0
12.5
12.5
25.0
SEEDS/GERMPLASM 15.0
15.0
ADMINISTRATION
130.0
15.0
91.0
13.0
11.0
TOTAL COSTS
505.0
200.0
116.0
55.5
133.5
Note: Some additional in-kind contribution will also be provided by GoK through the participation of
KEFRI
and
agricultural
extension
staff
involved
in
the
program.
27
PROPOSED ORGANIZATION OF THE KENYAN PROJECT
Steering group
ICRAF / KARI *)
Project lead / coordination
funding
GEF
funding
GoK
funding
WB
funding
Swiss Consortium
Representativ
of the Swiss
Novartis
ICRAF
field scientist
ICRAF
Lab scientist
ICRAF
social structure
KARI/KEFRI *)
SAirGroup
Roche
farmer association
Project
consultant
Swiss Re
Syngenta
NGO
(Note that the diagram identifies a consortium of Swiss multinational
companies for possible co-funding. This will be confirmed once the proposal has been
approved)
*) KARI and KEFRI are governmental
institutes
28
ANNEX 1 : METHODOLOGY FOR PROJECT SITE SELECTION
Overall rationale
Site selection for the project will begin at the highest level with the entire Kenyan portion of the Lake
Victoria basin, comprising all or parts of Western, Nyanza and Rift Valley Provinces. Justification for
this choice is based on the following criteria, as detailed in the main text:

Lake Victoria itself is suffering from high rates of sediment and nutrient loading from
unsustainable land use patterns in the watershed;

The watershed contains large areas of primarily agricultural, high potential lands with a
high probability of carbon gains from improved land use practices;

The high potential areas support high population densities, which are the areas with the
highest rates of degradation and with the greatest need for improved land use.
Selection criteria — from global to local
Within this broad region, potential sites will be scored based on the best available information
regarding several criteria, listed in order of global, regional and local importance. Failure to meet one
or another of the criteria will not exclude an area from consideration; rather an overall suitability score
will be developed to make selection decisions.
Data sources and methods to be used are detailed for each criterion. Where adequate data or
methodologies are not available, these will be developed and tested during the design phase of the
project. Criteria to be used are:

Potential for increased carbon sequestration through improved land use. Areas with high
rainfall and with finer textured soils, i.e. the higher potential areas, will be scored higher
on this criterion;

Potential for adverse impacts on water courses draining into Lake Victoria, based on
proximity to the lake or its major tributaries, and high risk for sediment and nutrient
transport to the lake;

Potential for enhanced landscape scale biodiversity through sustainable use;

Potential for biodiversity conservation in remnant areas of Western Rainforest and in
Montane Forest zones in the upper reaches of the watershed;

High levels of poverty, which represent areas where unsustainable agricultural practices
have accelerated land degradation.
Carbon sequestration potential criteria
The potential for C sequestration will be estimated based on the concepts of C deficit that has been
developed in the Alternatives to Slash and Burn Program and at ICRAF. C deficit is a means of
determining how much C has been lost from the landscape since clearing based on empirical
relationships that have been well validated. Estimates will be made for above- and below-ground C,
and then estimating the maximum amount that could be stored by the system under “best management
practices.” Since a large portion of the C that will be sequestered will be in the soil, much work can
29
be done through diffuse reflection spectrometry (DRS) and remote sensing. [Note: Because the
combination of DRS and remote sensing will be an important part of baseline development and of the
M&E system, details of these techniques will be presented in a technical Annex, Annex 2]
High potential areas will be targeted in this category, based on the following:

Rainfall > 1300 mm per year

Soil texture sandy clay loam or finer.
For rainfall, georeferenced databases exist which can delineate areas meeting the rainfall criterion. For
soil texture, the Exploratory Soil Map of Kenya (Sombroek et al, 1982) at 1:1,000,000 scale, is also
available in digital format. For the finer resolution needed for site selection, 1:100,000 scale maps are
available from Kenya Soil Survey in digital format for Siaya, Busia, Kisii and Vihiga districts.
Additional maps are available in paper format. At the finest resolution, both ICRAF and KARI have
extensive soil databases and sample archives, and most of the sample locations have been
georeferenced. It remains to convert these databases into GIS-based format.
Once broad delineations of soil factors have been made, final scoring will be based on a technique
developed by ICRAF scientists. The technique is based on diffuse reflectance spectra of soils, from
which most chemical and physical parameters of soils may be estimated rapidly and accurately (see
references by Shepherd and Walsh; Walsh and Shepherd; Walsh et al.).
Impacts on Lake Victoria criteria
Increased sediment and nutrient transport to Lake Victoria is in large part due to unsustainable land
use in the watershed. These practices include:

low- to zero-input agriculture, resulting in poor crop stands, accelerated erosion and
increased nutrient loss from the uplands (Rao et al., 1999);

Overgrazing in pastoral areas;

Unregulated timber harvest in the upper reaches of the watershed (in addition to the
biodiversity loss represented by these practices).
Identification of areas at greater risk for contributing to the lake’s pollution load will derive from
several sources of data. These will include comparison of historical and recent aerial photographs;
satellite imagery to identify sediment sources and outputs to the lake; and the newly developed soil
spectroscopy methods.
The DRS methods (see Annex 2), along with remote sensing data and digital elevation models, have
already been used successfully to develop an erosion risk map for part of the region. This work will be
extended to cover the entire Kenyan portion of the Lake Victoria Watershed. At this time, several
georeferenced transects along gradients of degradation in several sub-watersheds of the lake basin
have been completed. Diffuse reflectance spectra of these soils have enabled scientists to develop an
erosion risk index, and also its corollary, an index to rehabilitation potential. As an initial step, those
areas that have both a higher probability of degradation and a higher potential for rapid rehabilitation
will be targeted.
Proposed interventions in the at-risk areas of the watershed will depend on the particular problems
faced by each identified area that is contributing to degradation of the lake. These may range from
30
direct projects on improved land use, to policy studies aimed at strengthening institutional and
community controls on unsustainable timber harvest.
Biodiversity potential criteria
Given the significant predominance of smallholder farmer activities and their implications on the
potential for biodiversity loss and land degradation, a method will be developed for determining the
potential for increasing landscape biodiversity taking into account:

Sites with a significant percentage of smallholder producers amongst all producers;

Sites/ecosystems where the local institutional environment is conducive to introduction of
the proposed activities;

Areas without legal protection but with important biological distinctiveness, and where
land degradation process is starting as a result of small farmer activities;

Feasibility cost criteria.
Poverty and population criteria
High levels of poverty are generally associated with increased use of unsustainable agricultural
practices, and areas with high incidence of poverty will therefore be targeted for interventions. The
best poverty mapping data available are at District level, and are based on the 1989 census data. Data
from the 1999 census have not been released yet, but may be available within the PDF B timeframe.
Another source of data is the poverty-mapping project underway through collaboration between the
Kenya Central Bureau of Statistics and the International Livestock Research Institute that will be
completed. Results released by end 2001 and will be available for planning purposes for this project.
31
ANNEX 2: A SUMMARY ON DIFFUSE REFLECTANCE SPECTROMETRY (DRS)
Because DRS measurements will be central to the strategy to analyze landscapes to establish project
baselines and to monitor and evaluate project accomplishments, and because this technique is
relatively new, it is worth briefly reviewing the basis for its interpretation.
Many components of complex material mixtures (such as those contained in a soil sample) can be
distinguished using of their spectral signatures in the solar reflective region. Spectral signatures of
materials are defined by their reflectance or absorbance of light as a function of wavelength. Under
controlled conditions, the signatures are due to electron state transitions in atoms and vibrational
stretching and bending of groups of atoms that form molecules and crystals. Fundamental features (or
modes) in reflectance spectra occur at energy levels that allow molecules to rise to higher vibrational
states. The fundamental features related to various components of soil organic matter, for example,
generally occur in the mid- to thermal-infrared range (MIR, 2,500-25,000 nm), but their overtones (at
one half, one third, one fourth etc. of the wavelength of the fundamental feature) occur in the near(NIR, 700-1,000 nm) and short wave infrared (SWIR, 1,000-2,500 nm) regions. Soil minerals such as
different clay types have very distinct spectral signatures in the SWIR because of strong absorption of
the overtones of SO42–, CO32– and OH– radicals and combinations of fundamental features, for
example, of H2O and CO2. The visible (VIS, 400-700 nm) region has been widely used for color
determinations in soil and geological applications as well as in the identification of iron oxides and
hydroxides. Because of the close relationships between soil molecular/physical chemistry and soil
reflectance it is possible to consolidate assessment and prediction numerous soil properties under one
measurement.
Indeed, recent research has demonstrated the ability of DRS to provide non-destructive, rapid
prediction of soil physical, chemical, and biological properties in the laboratory (Dalal and Henry,
1986; Coleman and Montgomery, 1987; Morra et al., 1991; Palmborg and Nordgren, 1993; Ben-Dor
and Banin, 1995; Wander and Traina, 1996, Janik, et al., 1998; Ben-Dor et al., 1999). DRS has also
been used in the field, for instance to determine soil organic matter content (e.g. Sudduth and
Hummel, 1993). ICRAF work has shown that many different soil properties related to soil
productivity and degradation may be reliably predicted using laboratory and field based DRS
approaches (see Tables 1-3, Shepherd and Walsh, submitted; Walsh and Shepherd, in prep.).
While there is a growing body of literature related to the application of DRS to soil science there has
been little focus on examining the potential of soil reflectance as an integrated indicator of specific
soil functions, such as those related to primary productivity and soil degradation. ICRAF research
indicates that this may indeed be a promising area for further research. For example, crop
performance trials in Eastern and Southern Africa have been analyzed, with very good correlations
between soil reflectance and soil productivity (see for example Figure 1., Shepherd and Walsh,
submitted).
ICRAF experience has shown that spectral signatures can be successfully used to detect different land
management systems on a same soil type and can be used to detect degradation. Figure 2 shows how
the spectral signatures of soils that are cultivated differ from uncultivated soils and how spectral
signatures differ between soils subjected to different types of erosion. Finally, to demonstrate the
feasibility of application of DRS and remote sensing techniques to landscape scale analysis, Figure 3
presents a recent landscape scale analysis of erosion in the Nyando River basin that drains into Lake
Victoria.
32
Table A3-1. Prediction success in an 18-year soil management experiment in Kenya (from Shepherd
and Walsh, submitted, not for citation).
Soil attribute
r2cal*
r2val*
SEP†
Min
Max
Exchangeable bases (cmolc kg-1 soil)
0.90
0.81
0.796
6.3
12.8
Light fraction OM‡ (g kg-1 soil)
0.89
0.78
0.288
0.8
8.2
Microbial biomass C (mg kg-1 soil)
0.90
0.80
11.8
40
133
Bean yield§ (Mg grain ha-1)
0.91
0.82
0.092
0.22
1.01
Maize yield§ (Mg grain ha-1)
0.88
0.77
0.535
1.65
5.39
*Coefficients of determination for observed versus fitted values for calibration (n=31) and full crossvalidation sample sets. †Standard error of prediction. SEP for light fraction soil organic matter (SOM)
is presented for loge transformed data. ‡Light plus medium Ludox fraction of organic matter >250 µm
size and <1.37 Mg m-3 density. §Long-term average grain yields. Maize (Zea Mays L.) and beans
(Phaseolus vulgaris L.) were grown once each year in rotation.
33
Table A3-2. Relationships between soil attributes and soil reflectance in soil management
experiments. Coefficients of determination (r2) are for observed versus expected values of soil
attributes (0–15 cm depth) predicted from soil reflectance spectra convolved to Landsat 5 bandpasses (from Shepherd and Walsh, submitted, not for citation).
Soil Attribute
Study
Method
n
r2
Total soil N (g kg-1)
LTSM#
GM†
31
0.66
Macroorganic matter (g kg )
LTSM
GM
31
0.70
21
37
Light fraction N (mg kg-1)
LTSM
GM
31
0.78
23
126
Medium fraction N (mg kg )
LTSM
GM
31
0.71
4
75
Heavy fraction N (mg kg-1)
LTSM
GM
31
0.71
9
31
Microbial C (mg kg-1)
LTSM
GM
31
0.70
40
133
Microbial N (mg kg )
LTSM
GM
31
0.74
8
24
NaOH organic P (mg kg-1)
STAF-1*
GM
16
0.68
155
199
NaOH organic P (mg kg )
STAF-2
GM
16
0.62
62
113
Resin inorganic P (mg kg-1)
STAF-1
GM
16
0.34
2.3
4.4
Resin inorganic P (mg kg )
STAF-2
GM
16
0.77
5.7
18.7
Light fraction P (mg kg-1)
STAF-1
GM
16
0.33
0.1
2.2
Light fraction P (mg kg )
STAF-2
GM
16
0.39
0.1
1.6
Macroorganic matter P (mg kg-1)
STAF-1
GM
16
0.26
0.7
4.4
Macroorganic matter P (mg kg-1)
STAF-2
GM
16
0.52
0.5
4.4
§
‡
-1
-1
-1
-1
-1
-1
-1
Min
1.4
Max
2.2
Soil C (g kg )
MLAF
CC
114
0.76
6
32
Soil nitrate (mg kg-1)
MLAF
BR&
114
0.63
0.01
16.5
Exchangeable K (cmolc kg )
MLAF
CC
116
0.65
0.04
0.94
Extractable P (mg kg-1)
MLAF
BR
116
0.82
1.3
72.5
-1
#
LTSM is a long-term soil fertility management experiment in Kenya
STAF-1&2 are agroforestry experiments in Kenya (1) Oxisol, (2) Alfisol
§
MLAF is a multilocation agroforestry trial from Southern Africa.
†
Graphical model (Edwards, 1995)
‡
Canonical correlation analysis
&
Breakpoint regression analysis
*
34
Table A3-3: Prediction of basic soil properties using partial least-squares (PLS) regression of NIR
soil reflectance in the Lake Victoria Basin of East Africa (from Walsh & Shepherd, in prep, not for
citation).
Soil attribute
T (1)
# Comp (2)
Range (3)
r2cal (4)
r2val (5)
SEP (6)
SER (7)
PH (water)
Soil texture (8)
Clay (%) (8)
Silt (%) (8)
Sand (%) (8)
CEC Clay (cmol kg-1 clay)
Sum of Exch. Bases (cmol kg-1)
Ca (cmol kg-1)
Mg (cmol kg-1)
K (cmol kg-1)
Na (cmol kg-1) (9)
Org. C (g kg-1)
Min. N (mg kg-1 d-1)
Ext. P (mg kg-1)
none
mix
none
none
none
sqrt
sqrt
sqrt
sqrt
ln(x+1)
none
ln
ln(x+3.8)
ln(x+1)
13
10
10
10
10
5
8
10
10
10
10
15
13
10
4.8 – 10
–
5.0 – 79
0.0 – 42
8.0 – 90
4.0 – 188
0.3 – 55
0.6 – 48
0.0 – 18
0.0 – 6.2
0.0 – 6.7
2.3 – 56
-2.8 – 45
0.0 – 328
0.72
0.78
0.79
0.66
0.77
0.81
0.87
0.91
0.84
0.56
0.99
0.79
0.64
0.63
0.70
0.73
0.74
0.56
0.76
0.78
0.87
0.88
0.79
0.52
0.81
0.71
0.53
0.60
0.41
–
7.8
7.2
9.9
0.80
0.54
0.45
0.32
0.16
0.84
0.22
0.43
0.60
0.07
–
5.2
4.0
3.0
0.10
0.03
0.02
0.01
0.01
0.08
0.07
0.03
(1) type of transformation (indicated by Box-Cox test).
(2) number of significant spectral components in model.
(3) data range in original (untransformed) units
(4) Coefficient of determination for calibration set (n = 434, unless indicated as otherwise).
(5) Coefficient of determination for validation set (n = 217, unless indicated as otherwise).
(6) Standard error of prediction(in transformed units where applicable).
(7) Indicative standard error of replication in lab. data (in transformed units where applicable).
(8) PLS2 mixture model applied
(9) n = 32, full holdout cross-validation applied.
35
12
Predicted (kg ha
-1
)
10
8
6
4
2
0
0
2
4
6
8
10
12
Observed (kg ha-1)
Figure A3-1. DRS prediction of bean yield from an 18-year old soil fertility management
experiment in Kenya (solid circles are the calibration set, open circles are cross validated
values, from Shepherd and Walsh, submitted).
Z-value transformed reflectance
2.0
1.5
1.0
non-cultivated reference
0.5
cultivated reference
hardset
sheet eroded
0.0
gully eroded
-0.5
-1.0
-1.5
300
800
1300
1800
2300
Wavelength (nm)
Figure A3-2. Reflectance signatures of physically degraded soils in Western Kenya, relative to
“intact” non-cultivated and cultivated reference soils. This figure demonstrates the ability of DRS to
distinguish between physically degraded soils and to distinguish between management practices on
soils.
36
Landsat Thematic Mapper image of the Nyando River Basin (yellow outline) in
Western Kenya.
Image processed to highlight Nyando River sediment plume in Lake Victoria.
Image processed to highlight local sediment source areas.
Figure A2-3. Sample image analysis showing erosion/sedimentation at the level of a 3,500 km2 size
river basin draining into Lake Victoria in Western Kenya. Top panel shows the original image. The
middle panel shows the image processed to characterize the sediment plume in Winam Gulf from the
Nyando River. The bottom panel shows the area around Winam Gulf that is the likely source of the
sediment in the lake.
37
REFERENCES
Agricultural Technology and Information Response Initiative (2000). ATIRI booklet. Kenya
Agricultural Research Institute, Nairobi.
Ben-Dor, E. and Banin, A. (1995). Near-infrared analysis (NIRA) as a method to simultaneously
evaluate spectral featureless constituents of soils. Soil Sci. 159: 259-270.
Ben-Dor, E., Irons, J. R. and Epema, G. F. (1999). Soil reflectance, in Remote Sensing for the Earth
Sciences (Rencz, A. N. ed.), John Wiley & Sons, New York, p. 111-188.
Cleaver, K.M. and G.A. Schreiber. 1994. Reversing the spiral: The population, agriculture and
environment nexus in sub-Saharan Africa. World Bank, Wash, D.C.
Coleman, T. L. and Montgomery, O. L. (1987). Soil moisture, organic matter and iron content effect
on the spectral characteristics of selected vertisols and alfisols in Alabama. Photogramm. Eng.
Remote Sens. 12: 1659-1663.
Dalal, R. C. and Henry R. J. (1986). Simultaneous determination of moisture, organic carbon and total
nitrogen by near infrared reflectance spectrophotometry. Soil Sci. Soc. Am. J. 50: 120-123.
Desaeger, J. and M.R. Rao. 1999. The root-know nematode problem in Sesbania fallows and scope
for managing it in western Kenya. Agroforestry Systems. 47:273-288.
Hoekstra, D. and J. D. Corbett. 1995. Sustainable agricultural growth for the highlands of East and
Central Africa: Prospects to 2020. Int. Food Policy Research Institute, Washington, D.C.
IPCC. 2000. Lanuse, Land Use Change, and Forestry. Cambridge University Press.
Izac, A.-M. 2000. What paradigm for linking poverty alleviation and natural resources mamagement.
Special report to the CGIAR.
Janik, L. J., Merry, R. H. and Skjemstad, J. O. (1998). Can mid infrared diffuse reflectance analysis
replace soil extractions ? Aust. J. Exp. Ag. 38: 681-696.
Legume Research Network Project (1999). Issue No.1. Kenya Agricultural Research Institute,
Nairobi.
Lake Victoria Environmental Management Project (1996). Project document. Nairobi
Ministry of Environment and Natural Resources. 2000. The Kenya National Biodiversity Strategy
and Action Plan.
Morra, M. J., Hall, M. H. and Freeborn, L. L. (1991). Carbon and nitrogen analysis of soil fractions
using near-infrared reflectance spectroscopy. Soil Sci. Soc. Am. J. 55: 288-291.
NPEP, Office of the President. 1998-draft . National Poverty Eradiaction Plan.
Palmborg, C. and Nordgren, A. (1993). Modelling microbial activity and biomass in forest soil with
substrate quality measured using near infrared reflectance spectroscopy. Soil Biol. Biochem.
25: 1713-1718.
Rao, M.R., G. Mwasambu, M.N. Mathuva, A.A.H. Khan and P.C. Smithson. 1999. Effects of
phosphorus recapitalization and agroforestry on soil, water and nutrient conservation in
phosphorus-deficient soils of western Kenya. East Afr. Agric. For. J. 65:37-53.
Shepherd, K.D. and Walsh, M.G. (1999a). Sensing soil quality: Evidence from Africa. Working
Paper. Nairobi: ICRAF.
Shepherd, K.D. and Walsh, M.G. (1999b). Diffuse reflectance spectrometry in the NARL long-term
experiment-preliminary results. A report to the Kenya Agriculture Research Institute.
October 1999. Nairobi: ICRAF.
38
Shepherd KD and Walsh MG (2000) Light reflectance provides rapid assessment of soil quality.
Working Paper. Nairobi: ICRAF
Shepherd KD and Walsh MG (in progress) Diffuse reflectance spectrometry for rapid measurement of
soil properties and functional capacity. To be submitted to Soil Science Society of America
Journal.
Shepherd KD and Walsh MG (in progress) Measurement of management-induced changes in soil
quality using diffuse reflectance spectrometry. To be submitted to Soil Science Society of
America Journal.
Sombroek WG, Braun HMH and van der Pouw BJA 1982 Exploratory soil map and agro-climatic
zone map of Kenya, 1980. Kenya Soil Survey, Nairobi.
Soil Management Project and Legume Research Network Project (2000). Participatory technology
development for soil management by smallholders in Kenya. Special publication. Kenya
Agricultural Research Institute
Walsh MG and Shepherd KD (in progress). Screening for soil degradation in the Lake Victoria Basin.
To be submitted to Journal of Environmental. Quality.
Walsh MG and Shepherd KD (in progress). Remote sensing of soil physical degradation in the Lake
Victoria Basin. To be submitted to Remote Sensing of Environment.
Walsh M, Shepherd K, Verchot L, Albrect A, Noordvijk M van and Palm C (in progress). A rapid
method for assessing soil carbon saturation deficit. To be submitted to Soil Science Society of
America Journal.
Wamae, DK., Okoth P., Obanyi S. and Nandwa SM (1999). The potential use of geographical
information systems (GIS) as a tool to delineate agricultural inputs recommendation domains:
a case study of phosphate fertiliser application in Kisii district, Kenya. A paper presented at
the National Land and Water Management Conference 15th – 18th November 1999, Kenya
Agricultural Research Institute Headquarters, Nairobi.
Wander, M. M. and Traina, S. J. (1996). Organic matter fractions from organically and conventionally
managed soils: II. Characterization of composition. Soil Sci. Soc. Am. J. 60: 1087-1094.
Wass P (ed) 1995 Kenya’s Indigenous Forests: Status, Management and Conservation. ODA-IUCN,
Cambridge, UK.
White F 1983 The Vegetation of Africa. Natural Resources Research 20. UNESCO, Paris.
39